Reference Index — Sector 02: Rare Earths, Magnets, Qualified Supply Outside China

Companion reference file for the thesis (Parts III and VI principally). Every entry is researched fact with a working link. Market caps carry an as-of date. Compiled 2026-07-14.

MP Materials

What it is: The only integrated US rare-earth miner-separator-magnet maker (Mountain Pass mine → Fort Worth "Independence" magnet plant), and the thesis's central existence proof that the state can underwrite qualified supply outside China — capital in, a price floor, a procurement ban. Status: Public (NYSE: MP). The US Department of Defense is its largest shareholder via $400M of Series A convertible perpetual preferred (Jul 2025) plus a warrant for 15% of common at $30.03/share. Market cap / valuation: ~$9.29B as of 2026-07-10 (178.02M shares outstanding; stock +73.9% over the trailing 52 weeks). Latest financials: Q1 2026 (quarter ended Mar 31, 2026): revenue $90.6M, up 49% y/y; consolidated revenue plus PPA income $132.9M; Magnetics segment revenue $21.1M (up $15.9M y/y) and $9.6M segment adjusted EBITDA; net loss persists. Capex $77.4M in the quarter against FY2026 guidance of $500–600M as 10X construction accelerates. NdFeB magnet manufacturing began at Independence in December 2025; initial magnet revenue still guided to H2 2026. Cited for: The full DoD deal structure — $400M DoD Series A convertible perpetual preferred (+ up to $350M additional preferred committed), a $150M DoD loan for heavy-rare-earth separation (≈$550M of state money), a warrant, a 10-year NdPr price floor at $110/kg paid quarterly as the shortfall between $110/kg and the benchmark quarterly average VWAP (i.e. a contract-for-difference, ~2x China spot), and a 10-year DoD offtake covering 100% of 10X-facility magnet output; the $1B JPMorgan/Goldman financing; Apple's $500M offtake for recycled magnets from Fort Worth starting 2027; capacity going from ~3,000 MT/yr (Independence) to ~10,000 MT/yr with the 10X campus (Northlake, TX). Thesis: III.4 (the template); III.6 ("state can backstop qualified Western supply" — the live antidote to Molycorp); VI.5/VI.7 and VI OUTLINE §§64, 114 (the template for state-backed demand); III.5 (MP added to China's export-control entity list, Jun 2026). Note the OUTLINE's own instruction: use the disaggregated numbers, not the "$7.3B" headline, which is post-deal market cap. Links:

Lynas Rare Earths

What it is: The largest rare-earth separator outside China (Mt Weld mine, WA → Kuantan, Malaysia plant, plus Kalgoorlie), and the only commercial producer of *separated heavy* rare earths (dysprosium, terbium) outside China — the thesis's proof that the hard case, heavy REs, is buildable ex-China. Status: Public (ASX: LYC; ADR OTC: LYSDY / LYSCF). Independent; the Japanese state (JOGMEC/JARE) is a long-standing offtake and finance partner. Market cap / valuation: ~A$16.35B as of 2026-07-14 (share price A$16.24; +109.5% over the trailing year). Latest financials: Q3 FY26 (quarter ended 31 Mar 2026): record quarterly revenue A$265.0M, +115% y/y; sales receipts A$234.0M; closing cash A$1,070M. Total REO production 3,233 t (+69% y/y); NdPr production 1,996 t (vs 1,509 t prior-year quarter). First samarium oxide production in March 2026, one month ahead of schedule (~400 t/yr expected). Cited for: "First heavy-rare-earth producer outside China" — dysprosium oxide first produced May 2025 in Malaysia; 9 t of heavy rare earths (Dy + Tb) in the Jul–Sep 2025 quarter with first customer shipments; samarium oxide ahead of schedule (H1 2026); heavy-separation expansion; the Texas refinery (~US$575M, flagged with significant uncertainty and dependent on US government funding — as of 2026 Lynas has prioritised expanding Malaysian capacity instead). Also the $30.4M DoD Title III technology-investment agreement (light-RE separation, Texas) cited in VI. And Japan's post-2010 diversification-to-Lynas story (III.3.2). Thesis: III.4; III.6 (unlock: heavy REs outside China are buildable); VI ARGUMENT §273–275. NEW as of Q3 FY26 and worth folding in: the US has extended MP-style terms to Lynas — US$96M redirected to purchase Lynas product over four years with a US$110/kg NdPr floor price, plus a 12-year JARE agreement for 5,000 t/yr NdPr (expandable to 7,200 t). Links:

Noveon Magnetics

What it is: The only operational manufacturer of *sintered* NdFeB rare-earth magnets in the United States — the first company to reshore full-scale sintered magnet production in over two decades — and one of the three US magnet suppliers the thesis names as evidence the state template is working. Status: Private. Headquartered in San Marcos, Texas (150,000 sq ft plant on a 10-acre site within a 100-acre holding acquired in 2020). Proprietary "EcoFlux" closed-loop process using recycled feed. Market cap / valuation: Post-money valuation not disclosed. Latest round: $215M Series C, closed 19 Jan 2026, led by a $200M investment from One Investment Management (OneIM). Latest financials: Private; no revenue or margin disclosed. The operational figure it discloses is capacity: expanding beyond 2,000 tonnes/yr of finished sintered NdFeB, funded by the Series C, into automotive, defense, AI, energy and advanced-manufacturing demand. Cited for: "Sole US sintered-NdFeB producer; $215M Series C Jan 2026 (led by OneIM $200M) to expand beyond 2,000 tonnes/yr." Thesis: III.4 (western magnet plays standing up); VI.7 / VI ARGUMENT §282, §482 ("Noveon $215M Series C") as part of the magnet layer having the full state template. Links:

Vulcan Elements

What it is: A de-novo (2023-founded) US NdFeB magnet maker in Durham / Research Triangle Park, NC — the clearest case of the MP template being copied onto a startup: private equity, a very large state loan, government warrants, and a scale-up target measured in tonnes not units. Status: Private. Founded 2023 by John Maslin and Piotr Kulik. Vertically integrated with ReElement Technologies (oxides → metal → magnets), which has its own $80M OSC loan. Market cap / valuation: ~$2B, up from ~$200M, following the November 2025 federal package (per Wikipedia's summary of the reporting). Disclosed capital: $65M Series A (Aug 2025) from Altimeter Capital, One Investment Management and 1789 Capital; then a $1.4B package (Nov 2025) — a $620M Office of Strategic Capital direct loan (the largest OSC has ever made), $50M in Department of Commerce CHIPS incentives taken as equity, and $550M of private capital; the Department of War took warrants in both Vulcan and ReElement. Latest financials: Private; no revenue disclosed. Operating capacity was ~10 metric tonnes/yr as of October 2025; the funded plan is to build, commission and operate a 10,000 metric tonne/yr magnet facility, weighted toward recycled end-of-life magnets and e-waste feed. Cited for: "Vulcan Elements (Durham, NC) — first facility Mar 31 2025, ~10 t/yr by Oct 2025, $620M OSC loan (Nov 2025) plus a $65M Series A"; the SASC FY2026 defense bill authorizing $4M to buy Durham-made magnets. Thesis: III.4; VI.7 (VI OUTLINE §115; VI ARGUMENT §282, §515–516, §482). Note: the thesis's III.4 line "opened facility Mar 2026" conflicts with VI's "first facility Mar 31 2025" — the March 31, 2025 opening is the sourced date; the November 2025 event was a grand opening/expansion announcement at RTP. Links:

Niron Magnetics

What it is: Minneapolis-based developer of iron-nitride ("Clean Earth Magnet") permanent magnets containing no rare earths at all — the thesis's leading "design the magnet out" escape route from the chokepoint. Status: Private. Spun out of University of Minnesota research; backed by Samsung Ventures, Allison Ventures (Allison Transmission), Magna, GM Ventures, Stellantis Ventures, Volvo Cars Tech Fund, Anzu Partners, the Shakopee Mdewakanton Sioux Community and the University of Minnesota. Market cap / valuation: Post-money valuation not disclosed. Total funding ~$214M across 12 rounds from 19 investors (Tracxn profile, 2026). Disclosed rounds include $33M (Nov 2023, GM Ventures + Stellantis Ventures) and $25M (Feb 2024, led by Samsung Ventures with Allison and Magna). Latest financials: Private, pre-commercial-scale; no revenue disclosed. Non-dilutive support: a $52M award (Jan 2025) to advance US rare-earth-free magnet manufacturing (DOE 48C advanced-energy tax credit allocation, alongside earlier ARPA-E and DOE AMMTO support), plus a $10M Minnesota Forward Fund grant (Oct 2025). Plant: broke ground on a 1,500 t/yr, 190,000 sq ft facility in Sartell, MN — the world's first full-scale iron-nitride magnet production site — with operations anticipated to begin in early 2027. Cited for: "Niron Magnetics (iron-nitride / 'Clean Earth' magnets; ~$214M total funding; first commercial plant groundbreaking in Sartell, MN; ARPA-E + DOE AMMTO support incl. $52M award; automaker backers Samsung Ventures, Allison Transmission, Magna; Stellantis partnership)." Thesis: III.4 (magnet-free / reduced-magnet motors as escape route). Caveat the thesis should carry: Sartell's 1,500 t/yr is roughly one-seventh of MP's planned 10,000 t/yr, and iron-nitride magnets are not yet qualified in a robot joint. Links:

USA Rare Earth

What it is: A vertically-integrating US rare-earth-to-magnet company (Round Top deposit in Texas → metal/alloy via Less Common Metals → sintered NdFeB magnets at Stillwater, Oklahoma) — the third US magnet supplier outside China the thesis names, and, with MP, one of the two firms China blacklisted in June 2026. Status: Public (NASDAQ: USAR); HQ Stillwater, Oklahoma. Owns Less Common Metals (UK metal/alloy maker, acquired Nov 2025). Market cap / valuation: ~$4.16B as of 2026-07-14 (share price ~$18.20). Latest financials: Q1 2026 (quarter ended Mar 31, 2026): revenue $5.7M (all from Less Common Metals, skewed to Europe), gross margin 1.9%; loss from operations $36.7M; net loss $67.0M ( -$0.34/share). Balance sheet transformed by a $1.5B PIPE, taking cash to $1.75B and total assets to $2.13B. Operationally: commissioned Phase 1a of the Stillwater magnet plant, targeting 600 MTPA by Q4 2026; expanding LCM metal/alloy capacity toward 3,000 MTPA; first commercial yttrium metal production; signed a definitive ~$2.8B deal to acquire Serra Verde (Brazilian ionic-clay heavy REs); flagged potential access to $1.6B of CHIPS funding plus a $14.2M Texas grant. Cited for: "USA Rare Earth — magnet play; named on China's Jun 2026 entity list." On 22 June 2026 MOFCOM added 10 US entities — including MP Materials and USA Rare Earth, the two US-government-backed rare-earth firms — to its export-control list, barring exports of dual-use items to them with immediate effect, in response to the US expanding its own military-end-user list. Thesis: III.4 (western magnet plays); III.5 (the escalation calendar). Links:

Neo Performance Materials

What it is: The Toronto-listed rare-earth processor and magnet maker that runs the West's only *integrated* oxide-to-magnet chain outside China — separation at Sillamäe, Estonia feeding the new sintered-NdFeB plant at Narva — i.e. the European counterpart to MP, and the datapoint that a western magnet line can already ship samples to EV-motor customers. Status: Public (TSX: NEO; OTC: NOPMF). HQ Toronto; operations in Estonia, Thailand, China, Germany. Market cap / valuation: CAD ~1.92B as of 2026-07-14 (up ~229% over the trailing year). Latest financials: Q1 2026 (quarter ended Mar 31, 2026): revenue $155.0M (vs $121.6M in Q1 2025); adjusted EBITDA $36.2M (vs $17.1M); EPS $0.36, well ahead of consensus; FY2026 adjusted-EBITDA guidance raised to $100–110M from $75–80M. Cited for: Not named in the current thesis text — supplied as sector context for III.4's "western magnet plays now standing up." The relevant facts: the Narva, Estonia sintered-NdFeB plant opened 19 Sep 2025, the first integrated (oxide-to-magnet) NdFeB capacity commissioned in Europe, with initial capacity 2,000 t/yr scaling to 5,000 t/yr (Phase 1 ramp targeted by Q2 2026) and first magnet samples shipped to an EV-motor customer. It is a useful correction to any "MP is the only western magnet line" framing, and it belongs in the western-capacity arithmetic the OUTLINE asks for (MP ~10k + Noveon 2k + Vulcan + Neo 2–5k). Links:

Energy Fuels

What it is: A US uranium producer whose White Mesa Mill (Blanding, Utah) has become the only US site producing separated *heavy* rare earths — dysprosium and terbium, exactly the elements China put under licence in April 2025 — making it the domestic analogue of Lynas's heavy-separation claim. Status: Public (NYSE American: UUUU; TSX: EFR). Market cap / valuation: ~$3.27B as of 2026-07-14 (share price ~$13.25; +171.8% over the trailing year). Latest financials: Q1 2026 (quarter ended Mar 31, 2026): total revenue $35.8M (vs $16.9M Q1 2025), essentially all uranium — $35.7M from 510,000 lb of U3O8 at a realized $70.04/lb. No rare-earth revenue in the quarter. Net loss $10.8M ($0.04/sh), improved from $26.3M. Working capital $956.6M ($108.4M cash + $802.2M marketable securities). Cited for: Not named in the current thesis text — supplied as sector context for III.4/III.6 ("is any of this heavy-rare-earth (Dy/Tb) capable?", the OUTLINE's own open question). The facts that answer it: first kg of 99.9%-purity dysprosium oxide produced at White Mesa, Aug 2025; ~29 kg of Dy oxide through 2025; first US primary terbium oxide in decades, 99.9% purity, March 2026 (pilot scale, ~1 kg/week); US-produced heavy REO qualified for use in permanent magnets (Dec 2025). Commercial-scale Dy/Tb capacity is targeted from Q4 2026, and the Phase 2 circuit is scoped at ~288 t/yr Dy and ~80 t/yr Tb — but that remains guidance pending a final investment decision, so the heavy-RE gap is *demonstrated closable, not closed*. Links:

Cyclic Materials

What it is: Canadian magnet-to-magnet recycler (hub-and-spoke model: "Spoke" plants shred end-of-life magnets and e-waste into a rare-earth concentrate; a "Hub" refines it to oxides) — one of the two recycling escape routes the thesis names, and the route that reaches *heavy* rare earths without a mine. Status: Private. HQ Kingston, Ontario (Center of Excellence); first US Spoke in Mesa, Arizona. Investors include ArcTern Ventures, BMW i Ventures, Amazon's Climate Pledge Fund, Hitachi Ventures, Microsoft's Climate Innovation Fund, Jaguar Land Rover's InMotion Ventures. Market cap / valuation: Post-money valuation not disclosed. Latest round: US$75M Series C, closed 23 Jan 2026 (oversubscribed, its largest); total equity funding now above US$162M. Latest financials: Private; no revenue disclosed. Operating footprint: the Mesa, Arizona Spoke — described as the first commercial-scale scale-up for recycling and domestic production of rare earths in the US, covering both heavy and light rare-earth magnets — announced spring 2025; plus a recycling agreement with magnet manufacturer VACUUMSCHMELZE. Cited for: "Recycling: Cyclic Materials (magnet-to-magnet recycling) … western recycled-feed routes." Thesis: III.4 (recycling escape route), and it connects to the Apple/MP $500M offtake, which is explicitly for *recycled* magnets. The load-bearing point for the thesis is that recycled feed is the only western route that yields Dy/Tb without a new mine — the OUTLINE's open question about heavy-RE capability. Links:

REEtec

What it is: Norwegian rare-earth *separation* company building the first industrial REE separation plant in Europe outside the Estonian chain — the European mid-stream link that a non-Chinese magnet supply chain needs, and the second recycling and non-Chinese supply route the thesis names. Status: Private. Largest shareholder is LKAB (Swedish state-owned mining group, NOK 400M invested); other shareholders include Nysnø Klimainvesteringer (Norwegian state climate fund), Mercuria, TechMet and Scatec Innovation. Plant at Herøya, Porsgrunn; HQ Oslo. Market cap / valuation: No post-money valuation disclosed. Capital raised: NOK 1.2B (~EUR 115M) to fund the first industrial plant, the round in which LKAB became principal shareholder. Latest financials: Private; no revenue disclosed. Operational status: an industrial-scale demonstration facility has run since 2019; the first Herøya industrial plant is designed for ~720 t/yr of NdPr oxide, roughly 5% of estimated EU demand, with commissioning begun and raw-material feed from 2025; a second Herøya plant is planned, fed partly by LKAB's mining, targeted from 2027. REEtec has signed an REO offtake with Vital Metals. Cited for: "REEtec (Norway, separation) — western recycled-feed routes." Thesis: III.4. Honest caveat for the thesis: 720 t/yr NdPr oxide is a small fraction of the ~10,000 t/yr of *finished magnet* capacity MP alone is building, so REEtec is a proof-of-concept for European midstream, not a volume answer. Could not source a 2026-dated production update to a primary link as of 2026-07-14 — REEtec's own site still describes the plant as under construction/commissioning with LKAB as principal shareholder, and the most recent primary statements date to the NOK 1.2B raise and the LKAB ownership announcement. Links:

Molycorp (historical — 2015 Chapter 11)

What it is: The previous American attempt at building rare-earth supply outside China — the company that reopened Mountain Pass on the back of the 2010–11 price spike, spent $1.25B rebuilding it, and went bankrupt when Chinese prices came back down. In the thesis it is *the* failure mode the MP structure is designed to prevent. Status: Defunct as a public company. Filed Chapter 11 on 25 June 2015 in the US Bankruptcy Court for the District of Delaware (Molycorp, Inc. plus domestic and Canadian/Barbados/Luxembourg subsidiaries). The estate split: the Mountain Pass mine was sold and eventually became today's MP Materials; the downstream magnetics/Magnequench business was taken by creditor Oaktree Capital Management and reorganized as Neo Performance Materials (TSX: NEO, above). Market cap / valuation: At the 2010 IPO peak Molycorp was worth roughly $6B; by the filing the equity was worthless. The figure that matters is the liability side: it filed to restructure ~$1.7B of debt, after Oaktree won the bidding in 2014 to provide up to $400M of senior restructuring finance. Latest financials: Terminal. Its final SEC filings are the Chapter 11 monthly operating reports (8-Ks) and a Q2 2015 10-Q filed in bankruptcy; the company never reached sustained positive operating cash flow at Mountain Pass. Cited for: The 2010 embargo aftermath and the price-crash bust — dysprosium oxide rose from ~$90/kg (early 2009) to >$2,300/kg (mid-2011), the spike pulled Western capital in, Molycorp built a $1.25B facility, prices fell, and Molycorp filed Chapter 11 in June 2015. Thesis: III.3.2 (the precedent), III.6, and the closing numbers list at ARGUMENT §593-594. The load-bearing inference: a Western alternative supplier dies of *price*, not of engineering — which is why a state price floor (MP's $110/kg NdPr CFD) is the deliberate antidote and why the thesis treats the floor, not the plant, as the innovation. Links:

ZF Friedrichshafen (I2SM magnet-free motor)

What it is: German tier-1 driveline supplier whose I2SM (In-Rotor Inductive-Excited Synchronous Motor) is the most credible production-intent *magnet-free* traction motor — the thesis's second "design the magnet out" escape route, alongside Niron. Status: Private (stiftung-controlled): ZF Friedrichshafen AG is majority-owned (~93.8%) by the Zeppelin Foundation, administered by the city of Friedrichshafen. Not listed; it issues bonds, hence public financials. Market cap / valuation: No equity market cap — unlisted foundation-owned. The scale figure to use instead: FY2025 sales €38.8B with net debt €10.2B (as of 31 Dec 2025). Latest financials: FY2025 (year ended 31 Dec 2025): sales €38.8B (2024: €41.4B; organic growth +0.6%); adjusted EBIT €1.7B, margin 4.5% (2024: 3.5%); adjusted free cash flow €1.4B; 153,153 employees (2024: 161,631). Sold the ADAS business to Harman at an enterprise value of €1.5B. Cited for: "ZF I2SM production-ready inductively-excited magnet-free motor (CLEPA award; same torque/power density as PMSM, claimed)." Thesis: III.4 (magnet-free motors). The sourced specifics: ZF presented I2SM in late summer 2023 as a production-ready variant of a separately-excited synchronous motor, transmitting rotor excitation energy inductively *inside the rotor shaft* (no slip rings), with ~15% lower rotor energy-transmission losses than a conventional SESM; it won the CLEPA Innovation Award (Green category), announced in Brussels 4 Dec 2024. Important caveat the thesis should carry: ZF says it intends to bring I2SM to production maturity as an option in its e-drive platform — as of 2026-07-14 no confirmed series-production start date or named vehicle programme could be sourced to a primary link, so "production-ready" is ZF's characterisation of the design, not evidence of volume production. That matters because III.4's escape-route argument depends on whether the magnet can actually be designed out at volume, and it also connects to the ARGUMENT §148 point that a magnet-free *robot* joint means a bigger, heavier motor for the same joint torque. Links:

Proterial (ex-Hitachi Metals)

What it is: The Japanese materials group behind NEOMAX NdFeB magnets — the original Sumitomo/Hitachi NdFeB patent lineage — and the non-Chinese incumbent magnet maker that the thesis's "~200 NdFeB producers, top five ~17%" fragmentation claim implicitly includes; the Japanese half of the post-2010 diversification story. Status: Private. Formerly Hitachi Metals, Ltd.; acquired by a Bain Capital-led consortium (with JIC Capital and Japan Industrial Partners) and renamed Proterial, Ltd. on 4 January 2023. Delisted from the Tokyo Stock Exchange. Market cap / valuation: No listed market cap (taken private). Transaction value: the Bain-led tender offer was ~$5.7B for the equity, with the consortium deal reported at up to ~$7.5B including debt (2021 agreement, completed 2023). Latest financials: FY ended 31 March 2025: revenue ¥768.6B (~$5.2B). Product lines include NEOMAX NdFeB and ferrite magnets, amorphous metals, magnet wire and magnetic wedges; the company is publicly pursuing NdFeB magnet manufacturing in India as a non-China production base. Cited for: Not named in the current thesis text — supplied as sector context for III.4's fragmentation and alternative-supply arithmetic (the OUTLINE's "~200 NdFeB producers, top five ~17%" and the III.3.2 Japanese-diversification passage). Proterial is the reason the "three-company magnet monopoly" claim is wrong — a correction the thesis already carries as dead number #3 in III OUTLINE §43. Its ferrite-for-EV-motor push is also a *reduced*-magnet (rare-earth-free) route parallel to Niron and ZF. Links:

JL Mag Rare-Earth ("Jinli Permanent Magnet")

What it is: The world's largest producer of high-performance NdFeB permanent magnets — the single company most exposed to (and most enabling of) the humanoid magnet supply chain, supplier to Tesla, BYD and Toyota, and one of the three Chinese magnet makers granted MOFCOM's first general export licences in Dec 2025. Status: Public (SZSE: 300748; also HKEX-listed as 6680.HK since 2024). Ganzhou, Jiangxi. Market cap / valuation: CNY ~33.5B as of 2026-07-14 (Shenzhen A-share line). Latest financials: FY2025 revenue CNY 7.72B, +14.11% y/y, with earnings up 142.4% y/y. Production: 29,300 t of magnet blanks in 2024, with a stated plan to expand to 60,000 t/yr by 2027. It is piloting magnet assemblies for humanoid-robot joint motors and rotor components. Cited for: "Dec 2025: MOFCOM issued its first batch of general licenses (multi-shipment, ~1-year approvals) to major Chinese magnet makers — Jinli Permanent Magnet, Zhongke Sanhuan, Ningbo Yunsheng." Thesis: III.5 (the escalation calendar). JL Mag is Jinli. The point it carries: the licensing regime is not an embargo but a *throttle* — Beijing hands out general licences to its own champions, which is precisely the "capability to throttle magnet shipments through a licensing regime" described in III ARGUMENT §90. Its 29,300 t of 2024 blanks against MP's planned ~10,000 t/yr of finished magnets is the cleanest single-number scale comparison available for the western-capacity arithmetic the OUTLINE asks for. Links:

China Northern Rare Earth (Group) High-Tech

What it is: The world's largest rare-earth producer by volume, sitting on privileged access to the Bayan Obo deposit in Inner Mongolia — the physical basis of the country-level chokepoint the thesis argues for (the chokepoint is a *country*, not a cartel of magnet firms). Status: Public (SSE: 600111). Controlled by Baotou Iron and Steel (Group) Co., Ltd., a state-owned enterprise — i.e. ultimately a Chinese state asset, which is why it is the counterparty to the export-quota system rather than merely a participant in it. Market cap / valuation: CNY ~143.99B as of 2026-07-14 (+58.2% over the trailing year). Latest financials: FY2025: revenue ¥42.56B, +29.11% y/y; net profit attributable to shareholders ¥2.25B, +124.17% y/y; 2026 targets of >¥44B revenue and >¥3.5B profit. Q1 2026: revenue ¥11.859B, +27.69% y/y; net profit ¥918M, +113.12% y/y, on higher prices. Capacity: its Baotou Humei base, operational since October 2024, is the world's largest rare-earth feedstock facility at 106,661 t/yr of REO extraction and separation — against Bayan Obo reserves estimated above 35 Mt REO. Cited for: Not named in the current thesis text — supplied as the primary-source anchor for III's country-share claims: "~90% of rare-earth magnet processing, ~69% of mining" and "~200 NdFeB producers, top five ~17% → chokepoint at country level" (III OUTLINE §201–202). One state-controlled group with a single 106,661 t/yr separation base and the Bayan Obo reserve is what makes the *country* the chokepoint even though the magnet industry downstream is fragmented. Its surging 2025–26 margins on "higher prices" are also the direct financial fingerprint of the April-2025 licensing regime. Links:

Cross-cutting notes for the thesis

Western finished-magnet capacity, 2026–2028 (as sourced above): MP ~3,000 t/yr (Independence, producing since Dec 2025) → ~10,000 t/yr with the 10X campus; Noveon >2,000 t/yr; Neo (Narva, Estonia) 2,000 t/yr scaling to 5,000; Vulcan ~10 t/yr today, funded to 10,000; Niron (Sartell, rare-earth-free) 1,500 t/yr from early 2027. Against JL Mag alone at 29,300 t of blanks in 2024, heading for 60,000 t by 2027.
The single company most at risk of being over-relied on is MP: the thesis is right that the flagship is "one company deep." Neo (Estonia) and Energy Fuels (heavy REs) materially soften that, and both should be named.